Residential treatment is one of the most desirable modalities offered in any treatment system. Yet it is also one of the most costly, and its benefits are limited if not followed reliably by continuing aftercare in an outpatient program. This services research project will examine the comparative effectiveness of three evidence-based interventions for improving historically low rates of successful transition from residential to outpatient care. The project takes place within the treatment service delivery system of Baltimore City and represents a collaboration between Johns Hopkins University SOM, the Mid Atlantic Node of the National Drug Abuse Treatment Clinical Trials Network (CTN) and Baltimore Substance Abuse Systems (BSAS), the Substance abuse authority for the City of Baltimore. The interventions to be examined are a) Transportation Escort that delivers residential clients to the door of the outpatient referral program on the day of discharge, b) Client Transition Incentive that offers a financial incentive to clients who successfully enroll in outpatient aftercare and c) Residential In-reach that arranges. orientation visits by outpatient staff who visit referral clients at the residential program. These enhanced interventions will be compared with a usual care control condition in which no special intervention is in place. Using a Latin Square cross-over design than maximizes external validity, treatment conditions will be in place during 6-week intervals of time in a pre-determined sequence during the 2-year project. Primary outcome measure will be the percent of referred clients who are admitted to an outpatient aftercare program within 30 days of residential discharge. Length of stay in the outpatient program will also be examined as a function of transition interventions as a secondary outcome. Overall, the project will provide valuable new data that can be used by policy makers to select treatment transition options that will maximize continuity of care within a drug abuse treatment services delivery system.